Professional Documents
Culture Documents
Class and Ocupation.: Wright, Erik Olin
Class and Ocupation.: Wright, Erik Olin
Cita:
Wright, Erik Olin (1980). Class and Ocupation. Theory and Society, 9 (1),
177-214.
ARK: https://n2t.net/ark:/13683/paqp/3qk
Acta Académica es un proyecto académico sin fines de lucro enmarcado en la iniciativa de acceso
abierto. Acta Académica fue creado para facilitar a investigadores de todo el mundo el compartir su
producción académica. Para crear un perfil gratuitamente o acceder a otros trabajos visite:
https://www.aacademica.org.
177
that labor within the labor process. Workers are thus definable only in
terms of their social relationship to capitalists, not in terms of the techni-
cal content of their laboring activity. A carpenter, for example, could be
located in any of a number of different class positions: worker-carpenters
are wage-laborers for capital who lack any significant control over their
labor process; petty bourgeois carpenters are self-employed artisans
who sell carpentry services directly to consumers; manager-carpentersare
wage laborers who control the labor of other carpenters within produc-
tion (e.g., foremen). Within marxist theory, therefore, it is impossible to
define classes as clusters of occupations; class and occupation occupy
basically different theoretical spaces.
This paper will explore the complex relationship between class and occu-
pation from a marxist perspective. While there will be periodic references
to the contrast with nonmarxist conceptions of class, the central objec-
tive is not to engage in a battle of definitions with nonmarxists, but to
clarify these concepts within marxist theory itself. The heart of the analy-
sis, therefore, will be abstract and conceptual. This may give the paper a
rather formal and schematic flavor. The underlying premise is that this
kind of conceptual clarification is essential if empirical and historical
research is to procede in a clear and coherent way.
The claim that classes are defined by the social relations of production
does not, by itself, provide an adequate basis for defining specific classes
within a specific social organization of production. The social relations of
production have many different aspects and can be characterized in many
different ways. Some more general principle for the decoding of the social
relations of production, therefore, is needed if we are to be able specify
the criteria for the classes which those relations determine.
(a) Social relations of control over money capital, i.e., control over the
flow of investments and the capital accumulation process.
(b) Social relations of control over physical capital, i.e., control over the
use of the physical means of production.
(c) Social relations of control over labor, i.e., control over supervision
and discipline within the labor process.
money as capital); (b) that they do not control physical capital, for if they
did they could use that physical capital to produce products for their own
use; (c) that they do not control the labor process itself, for if they did
they could reduce the pace of work to the point that no surplus labor
would be performed. Capitalist exploitation, and the class structure
which such relations of exploitation determine, is thus the specific conse-
quence of this combination of these three dimensions of social relations
of production.
| Bourgeoisie |. ..
JSmall employers'
r---J--- ?-_---- --j^^I
[ Managers and] ^ ' orei<
',supervisors i r_________ - P g3
-i--
. I Semni-autonomnous i
, ^^-"---wage-earners_ J
| Proletariat ~
I i Classes
OperationalizingClass Relations
Itis one thing to elaborate a
definition of classes; it is quite another to
developan adequate
areimmediately apparent. operationalization of that definition. Two problems
First,it should beclear from the above discus-
sionof contradictory class locations
that the precise boundary criteria
betweencontradictory locations and the
proper class locations of capi-
talistsociety are rather ambiguous.
example, how much "autonomy"
isnecessary to define a worker as For
semi-autonomous, as occupying a
contradictory location between the working class and the
sie? petty bourgeoi-
Surely the criterion of absolutely any
autonomy is too broad. While
historical data on the labor process are rather
meager, it is unlikely that
more than a small fraction of the
working class was ever characterized by
the classic
image of the fully proletarianized worker, totally under the
controlof the capitalist through a minutely subdividing labor
governedbyprinciples of scientific process
management. Most workers, most of
thetime, have been able to maintain at least
some residual control over
their immediate labor process. Similarly, it would
restrictthe concept of "semi-autonomy" to beinappropriate to
positions which, like univer-
sity
professors, have extremely high levels of control
over the pace of
work, the scheduling of work, the content
of work,
thatthere is no absolute criterion which etc. The difficulty is
defines how much control is
sufficientto exclude a position from the
proletariat. A certain arbitrari-
ness,therefore, will inevitably enter into any
attempt to measure this
semi-autonomous class location. Similar problems
boundaries of the other contradictory occur in specifying the
locations. While this does not
imply that it is impossible to
more operationalize such class locations - any
than it is impossible to define
"mammals" because of the existence
oftheplatypus - it does mean that any estimate of the
size of a given class
location will involve upper and lower bounds
rather than a single figure.
Apart
from the problem of the arbitrariness
of formal criteria for boun-
184
Bourgeoisie | \ r
1- 2%
L
i Smallemployers I
Top managers L 6-7%
Imiddle managers II
technocrats \
1L _11-17%
________
- - Pe t t y bo urg e o is ie
Bottom 6.0 ?/o
i ana3gers i
Iforemen and line I
Isupervisors I
I 13-23?%/ r--I- ....
-L __ T ... ..--- - Semi-autonomous I
I | / _I employees
Proletariat
41-54%
r 5-11% .
Classand Occupation
Thisuneven distribution of
occupations and classes is even more strik-
inglyrevealed if we look at the distribution of
classes within occupations:
71%of all unskilled laborers are in the
working class, whereas only 14%
ofall professionals are workers. In
contrast, only 17%of all laborers are
managers/supervisorscompared to 69% of all
professionals.
Thecommon association of the
working class with manual occupations,
andnonmanual occupations with the
"middle" class, is thus not a
totallyunrealistic picture. The process of
laborprocess has certainly moved proletarianization within the
much more rapidly among most
manual occupations than among white collar
occupations. Whether or
not
nonmanual occupations will undergo as
pervasive a process of prole-
tarianization
in the future remains to be but for the moment it is
that the heart of the working class seen,
clear
remains embedded in manual
occupations within the technical relations of
production.
However, it is important not to take this
observation too far. The data in
Table3B also indicate that there are
very different levels of proletariani-
zationamong white collar occupations and among
manual occupations.
Our
third general conclusion is that lower
white collar occupations are
muchmore like lower blue collar occupations in
their class distributions
than
they are like upper white collar
and that crafts blue
collaroccupations are much moreoccupations,
like technician white collar
occupationsthan they are like lower blue collar
of occupations. Nearly 60%
allpeople in clerical occupations and
collar
occupations fall into the working class, 64%of people in lower blue
compared to only 17%of
in upper white collar occupations.
people
Among both craftspersons and
on the other hand,
technicians, approximately 39% are workers; 12%of
all
technicians and 15%of craftspersons are
and of technicians and 38%of semi-autonomous employees;
45% craftspersonsare managers/ supervisors.
188
Class
and Organization
Oneof the fundamental distinctions within
classical marxist theory has
beenbetween a class-in-itself and a class-for-itself.
The former is gen-
erally
understood as structurally defined locations within
the social rela-
tions
of production. The latter, on the
other hand, implies classes
organized
as social forces, as collective actors in
history.20
This
distinction between a class-in-itself and
termsof the concept of organization. A for-itself can be recast in
classes class-in-itself can be defined as
within organizations; a class-for-itself as organizations within
classes.
The social relations of production which
define class positions are
locatedwithin organizations, principally
enterprises (i.e., capitalist eco-
nomicorganizations), but also political and ideological
state apparatuses). The "empty places"
(i.e., organizations
defined by those social rela-
tions
of production to use Adam
- Przeworski's apt expression are -
190
Firm A Firm B
Capital Capital
Labor Labor
/ \ Union |
Occupationsand Organizations
(1) Occupational positions form the potential basis for persistent divi-
sions within the working class. In spite of occupying a common location
within the social relations of production, significant internal divisions
exist within all working classes. There are a number of different structur-
al bases for these internal divisions: racial and ethnic divisions; employ-
ment in monopoly, competitive or state sectors of the economy;25
unionized versus ununionized employment; levels of skill, income, educa-
tion, status of different categories of workers. One of the central prob-
lems of a class analysis is to sort out the relative salience of these different
sources of fractionation of the working class and analyse the conditions
under which such divisions may be transcended.
(3) Occupational positions may not simply form a structural basis for
segments or strata within classes; they may also compete with class as the
basis for organizing the empty places within production. To the extent
that they are formed into collective organizations, occupational groups
can serve to obscure class issues, class divisions, class conflicts. The
classic example of this is the formation of farmers into collective organi-
zations which obscure the qualitatively different class situation of share-
croppers, small family farms, corporate farms, etc. Similarly, the for-
mation of all physicians into professional associations such as the
AMA obscures the quite different class situation of physicians in the state
sector, in large hospitals and in private practice. Such professional-oc-
cupational formation probably contributes to the general lack of support
for strikes by hospital workers and nurses on the part of physicians.
I will now try to take these general observations and restate them more
systematically in terms of a formal model of determination.
To briefly recapitulate the argument thus far: classes are defined by the
social relations of production, occupations by the technical relations of
production. Both of these relations are decisive dimensions of the internal
structure of capitalist organizations of production, and both of them
determine the empty places within production. In the process by which
classes within organizations (a class-in-itself) are transformed into
organizations within classes (a class-for-itself), the precise relationship
between class and occupation plays an important role, by creating
internal strata within classes, by potentially obscuring basic class
divisions, and, under certain conditions, by providing resources for class
conflict.
A v > C
In this case B determines the effects of A on C. In our analysis of class and
occupation, both limitation and selection determinations are themselves
196
(2) Class Structure Limits Class Formation. The social relations of pro-
duction also set the basic limits on the formations of organizations within
classes. For example, as long as capitalist social relations of production
were primarily organized in small shops and cottage industry, the forma-
tion of industrial trade unions was basically outside of the limits of
structural possibility. Or, to take a different sort of example, where there
is a large proportion of all wage laborers in contradictory locations
within class relations, it may be much more difficult to organize the
working class itself into a class party (either a revolutionary or reformist
party), since it will be more difficult for individual workers to see the class
structure as a polarized, antagonistic structure of class domination. To
the extent that the existence of contradictory'locations obscures class
interests, they may place real limits on the possibility of working class
formation at the political level.32
(4) Class Formation Limits Class Struggle. The basic limits on classstrug-
gle are not set directly by the class structure but rather by the forms of
organization of classes. Without certain forms of organization, certain
types of struggle are simply impossible, regardless of the underlying class
structure. Collective bargaining over economistic demands as a form of
struggle presupposes union organizations; electoral struggles over imme-
diate class interests presuppose the existence of working class parties.33
This is precisely why many of the bitterest forms of class struggle center on
the establishment and legitimation of different organizations of the work-
ing class. The struggle for the creation of unions, for example, is typically
much sharper and bloodier than the struggle over various objectives once
unions are in place. The introduction of new forms of organization of
classes transforms the terrain upon which subsequent struggles are fought
and thus often meets with the staunchest resistance by ruling classes.
(5) Class Structure Selects Class Struggle. Within the limits established
by the forms of organizations-within-class, the class structure, like the
occupational structure, acts as a selection determinant on forms of class
struggle. The extent to which a working class political party, for example,
organizes its activity around immediate or fundamental class interests
may be shaped by the relative size of the contradictory locations within
class relations among wage-laborers. Fundamental class interests would
tend to highlight the contradictory character of the potential alliances
between workers and people in contradictory class locations near the
working class, and thus make electoral gains more difficult. As Adam
199
(6) Class Struggle Mediations. Class struggle mediates all of the relations
of determination between the structural elements of the model. The conse-
quences of a given structure of production relations for the technical
relations of production or for class organization, therefore, cannot be
considered directly given by the class structure itself. Those consequences
- the ways in which the class structure limits the occupational structure
and class formation - are themselves shaped by the actual, historical
forms of struggle that exist within those structures. For example, if capi-
tal attempts to deal with economic crisis through a general assault on the
living standards of all wage-earners, this may increase the possibilities of
class alliances between the working class and contradictory class loca-
tions. If on the other hand, an attempt is made by the capitalist class to
concentrate the effects of crisis on the weakest segments of the working
class, it is possible that it could deepen the conflict of interest between
certain contradictory locations and workers. The strategies of capital -
its practices within the class struggle - thus mediate the effects of class
structure on working class formation. But equally, the strategies of the
working class within the class struggle mediate those effects. Where work-
ers organize their demands around fundamental class interests, interests
which call into question the capitalist mode of production itself, the
possibilities of a cohesive class organization of the dominant class increas-
es. When workers are either relatively passive or restrict their demands
to immediate interests, on the other hand, the various fractions of the
capitalist class are much more likely to be hostile to each other and
engaged in serious political and economic conflict. The effects of the
class structure on the class formation of the bourgeoisie is thus mediated
by the forms of struggle of the working class.
(7) Class Struggle Transformations. Finally, and in the end perhaps most
critically, class struggle acts as a transformation determinant directly
on each of the structural elements in the model. When marxists say that
class struggle is the "motor" of history, it is largely because of the trans-
formative role that class struggle plays in shaping and reshaping the
social structure itself. In every period of capitalist development, the deter-
mination of the very form of working class organizations has been a
200
decisive object of class struggle. This has included such things as struggles
over union rights, over freedom of political parties, and periodically over
new forms of class organization within production itself (workers' coun-
cils and the like). Transformations of the occupational structure have
also been objects of struggle. From the working class' point of view, such
struggles have often centered on preventing changes in the technical divi-
sion of labor which eliminate skilled jobs; from the capitalist class' point
of view, such struggles have centered on using technology and the occupa-
tional structure as part of a strategy of social control.36 In both cases,
the structure of occupations has been an object of struggle. Finally,
struggles over the class structure itself constitute in some sense the
ultimate objective of class conflict. A socialist revolution is above all a
revolutionary restructuring of the class structure. But struggles over the
class structure are not limited to revolutionary transformation: they also
comprise a central aspect of capitalist development itself. The classic
example of such struggle is the destruction of the petty bourgeoisie in the
course of capitalist development, first through such things as enclosures
which destroyed much of the agrarian petty bourgeoisie, and later
through capitalist competition which destroyed the artisanal petty bour-
geoisie and is now making considerable inroads on the retail trade petty
bourgeoisie as well. Perhaps more importantly in the present context,
class struggles have directly influenced the shape and size of the contra-
dictory class locations near the working class. On the one hand, the
expansion and differentiation of a managerial structure is at least in part a
response to the imperatives of domination within large scale capitalist
production (and not simply the imperatives of technical coordination).37
On the other hand, the need of capital to control the labor process has
probably led to a systematic attempt at reducing the degree of control
overthe labor process and thus to a contraction of the semi-autonomous
employee category.38
The overarching thesis of this essay has been that class structure cannot
be considered simply a typology of occupational categories, regardless of
how refined that typology attempts to be. Class and occupation are
different dimensions of social structure, and any serious attempt at
understanding the dynamics of social change must try to theorize the
relationship between these dimensions rather than collapsing one onto
the other. Once the distinction is rigorously made between class and
occupation, between the social relations of production and the technical
division of labor, we can then embark on the task of linking the two
within a general model of determination of class formation.
TABLE 1:
' Since 80% of all employers in the sample employed less than ten workers, it was not
possible to study a proper capitalist class location. Throughout most of the analysis
which follows, therefore, I will treat all employers as occupying a contradictory location
between the petty bourgeoisie and the capitalist class.
2 All teachers were classified as nonsupervisors regardless of their response to this
criterion (see text for explanation).
3 Jobs which the respondent claims are characterized "a lot" by both of the following
descriptions:
a) "a job that allows a lot of freedom as to how you do your work"
b) "a job that allows you to make a lot of decisions on your own".
203
TABLE 2:
Managers/Supervisors
Top/middle All supervisors who also Supervisorswith say in pay
managers reportthat they have andpromotions whose occupa-
some "say in the pay and tion is classified as profession-
promotions"of their al, technical, managerial, or
subordinates.2 official.
TABLE 3:
Managers, proprietors,
and officials 70.5 30.4 16.6 3.6 0.9 13.4 (206)
TABLE 3, continued
Managers, proprietors,
and officials 38.5 13.6 42.2 2.9 2.9 100.0
TABLE 4:
Managers, proprietors,
and officials 69.0 0.0 18.3 2.7 1.2 16.7 (166)
TABLE 4, continued
Managers, proprietors,
and officials 41.6 12.1 42.2 1.8 2.4 100.0
TABLE 5:
Managers, proprietors,
and officials 83.3 47.1 12.0 5.2 .6 7.4 (40)
Upper blue collar 0.0 0.0 0.7 0.0 0.6 0.6 (3)
TABLE 5, continued
Managers, proprietors,
and officials 25.0 20.0 42.5 7.5 5.0 100.0
NOTES
1. Frank Parkin, Class Inequality and the Political Order (New York:
Praeger, 1971),
p. 18.
2. For example, E. P. Thompson, The Making of the English Working Class (New York:
Random House, 1966).
3. For example, Nicos Poulantzas, Classes in Contemporary Capitalism (London: New
Left Books, 1975).
4. Martin Nicolaus, "Proletariat and Middle Class in Marx," Studies on the Left, no. 7,
1967; and John Urry, "Towards a Structural Theory of the Middle Class," Acta Socio-
logica, Vol. 16, no. 3, 1973.
5. Poulantzas, Classes in Contemporary Capitalism.
6. Erik Olin Wright, Class Structure and Income Inequalit,, unpublished Ph.D. disserta-
tion, University of California, Berkeley, 1976;"Class Boundaries in Advanced Capitalist
Societies," New Left Review, no. 98, pp. 3 -43; Class, Crisis, and the State (London:
New Left Books, 1978), ch. 2; Class Structure and Income Determination (New York:
Academic Press, 1979), ch. 1; "Alternative Perspectives on the Marxist Concept of
Class," Politics & Society. Vol. 9/3, 1979.
7."Surplus labor" is the general category, applicable to all class societies. It is only in a
capitalist society, however, that such labor is embodied in commodities as "exchange
values," and that the surplus labor represented in the surplus product takes the form of
"surplus value." Much of Marx's discussion of capitalism in Volume I of Capital is
devoted to explaining precisely how surplus labor takes the form of surplus value in
capitalism.
8. The technical defense of this account is at the heart of Marx's labor theory of value. The
basic claim that surplus labor is appropriated through the difference between the labor
embodied in the wage bundle and the labor embodied in the social product, however,
does not hinge on the general adequacy of the labor theory of value as a framework for
understandingthe rate of profit and other aspects of the accumulation process. All that
is necessary to defend the description of this mechanism as a mechanism of exploitation
isto demonstrate that the wage bundle in fact embodies a certain quantity of social labor
and that this is less than the social labor embodied in the total social product. Thevery
fact of the existence of a surplus product controlled by the capitalist class which was
produced by the labor of workers establishes this. Fora general discussion of the labor
theoryof value and its relationship to the analysis of class,see Wright, "Currentdebates
on the Labor Theory of Value," New Left Review (forthcoming).
9.Note that these are referred to as aspects of production relations, not types of relations.
Thereis no meaningful sense in which they canexist independently of each other. They
representthree intrinsically interlinked aspects of the social relations of production,
forming a unity, albeit an internally differentiated (or structured) unity.
10.Relational concepts of classand inequality should be contrasted to gradational notions.
In gradationalnotions, classesdiffer in the degree of various attributes such as income,
status,education, wealth. The names of classes thus reflect purely quantitative locations:
upper class, upper middle class, middle class, lower middle class, lower class,etc. In
relationalnotions, on the other hand, classes are defined byvirtue of their location
withinsocial relations of domination/subordination. The names of classes, therefore,
reflect the qualitative nature of the locations within such relations: lord and serf;
slavemasterand slave; capitalist and worker. Forextended discussions of the distinction
betweenrelational and gradational conceptions of class, see S. Ossowski, Class Struc-
turein theSocial Consciousness (London: Routledge & Kegan Paul, 1963), and Wright,
ClassStructure and Income Determination, Ch. 1.
11.Commercial airline pilots pose an additional problem for a classanalysis. Their incomes
may besufficiently high that they canno longer beseen as "exploited" in the convention-
211
al marxist sense of the term. In such cases there is a contradiction between their
location within production relations and their location within exchange relations. Such
nonexploited wage-laborers should not be considered part of the working class, even if
at the level of production relations they have no meaningful control over their labor
process. For the purposes of this paper this additional complication will be ignored, and
we will examine class locations as they are determined solely at the level of the social
relations of production.
12. The complex relationship of intellectuals to working class social movements reflects this
contradictory determination of their class location. For a discussion of the relationship
of intellectuals to the working class, see Wright, "Race, Class and Income Inequality,"
American Journal of Sociology, May, 1978.
13.1 am currently engaged in a large scale, cross-national study of class structure in the
United States, Italy, Sweden and Great Britain which will generate systematic data on
class relations directly based on the theoretical categories discussed above. Results from
this research should be available by late 1981.
14. The central issue here is that teachers (in most circumstances) are not engaged in the
exploitation of labor power. While they do control the activity of students, they do not
control the labor of direct producers (workers).
15. Since this chart is based on data limited to the economically active, working population,
it cannot be considered a complete class map of the American population. Unemployed
people, students, retired people, and nonworking housewives are not included in the
data. It is, of course, an important theoretical and empirical problem to understand the
relationship of such positions outside the market to class relations in capitalist society.
The data available for the present study, however, do not allow such questions to be
investigated. See Wright, Class Crisis and the State, 91 - 96, for a discussion of the class
location of these positions.
16. Since so-called "discouraged workers" as well as temporarily unemployed workers are
not included in the Survey of Working Conditions, and since most such people would
be drawn from the working class, these estimates almost certainly underestimatethe size
of the American working class, even if we assume that all of the other criteria are
adequate.
17. These results are quite consistent with the findings of Reeve Vanneman, "TheOccupation-
al Composition of American Classes," American Journal of Sociology, 82, 4, 1977,
pp. 783 - 808, in a statistical study of the occupational composition of "classes."Vanne-
man used cluster analysis to establish the closeness of different occupational categories
to each other in terms of two basic variables: residential segregation and intergeneral
mobility. While, as Vanneman admits, these clusters of occupations cannot be consid-
ered "classes" in analytical terms since they are purely statistical aggregations, never-
theless his results are very suggestive. He found that lower white vollar occupations are
much "closer"to traditional blue collar occupations in terms of both residential segrega-
tion and intergenerational mobility than they were to other white collar occupations. To
the extent that location within the social relations of production establishes limits of
variation on such things as intergenerational mobility, it would be expected that occupa-
tions with similar class distributions would have similar mobility patterns. Vanneman's
study supports such a proposition.
18. It is always possible, of course, that the differences between men and women in Tables 4
and 5 are a consequence of the level of aggregation of occupations. With more detailed
distinctions within broad occupational categories it could turn out that while men and
women have different occupational distributions, within a given occupational category
(position within the technical division of labor) they have the same class distribution.
The observation that occupational levels of proletarianization vary with sex, therefore,
must be seen as tentative until more elaborate data on class and occupation is available.
19. In some ways, "class struggle" is even more difficult to define rigorously than class
structure. Throughout this discussion I will use the term in its broadest possible sense to
212
include both social conflicts between self-consciously organized classes and conflicts
which are directly implicated in the formation and transformation of classes, even if the
"combatants" in the conflict are not strictly speaking class actors. To use Przeworski's
formulation, class struggle must be understood as involving both the struggle between
classes and the struggle over class. It should also be noted that throughout this discus-
sion I will use the terms class struggle and class conflict interchangeably. See Adam
Przeworski, "The Process of Class Formation: From Karl Kautsky's The Class Struggle
to Recent Debates," Politics and Society, 7, 4, 1977.
20. There has been considerable debate in recent years over this distinction. Traditionally,
marxists have tended to identify a class-in-itself with the economic location of classes,
and a class-for-itself with the political and ideological formation of classes. This has lead
some marxists, most notably Nicos Poulantzas to reject the distinction between in-itself
and for-itself altogether. As Poulantzas has correctly emphasized, even the structural
definition of classes requires political and ideological elements. the distinction class-in-
itself/for-itself, however, does not hinge on the distinction between political-ideological
relations and economic relations. Rather, the theoretical impulse for the distinction is to
distinguish classes as positions from classes as organized social forces. It is perfectly
consistent to sustain this distinction and to argue that class positions themselves need to
be analyzed in terms of political and ideological dimensions. For Poulantzas' position,
see his Political Power and Social Class (London; New Left Books, 1973) and Classes in
Contemporary Capitalism (London: New Left Books, 1975).
21. Stated in somewhat different terms, a class-in-itself is defined fundamentally by the social
relations between classes; class-for-itself by the social relations within a class. The social
relations between classes define the essential structure within capitalist enterprises; the
social relations within classes define the essential structures of the class organizations
across enterprises.
22. Goran Therbon develops a notion of organizations which is very similar to the one
argued here, in the context of an analysis of the capitalist state. He writes: "In order to
understand the class character of the state apparatus, then, we must begin to develop a
new approach to the study of organization. We should view it not as a goal-oriented
subject in an environment but as a formally bounded system of structured processes
within a global system of societal processes." See What Does the Ruling Class Do
When It Rules? (London: New Left Books, 1978), p. 37.
23. Note that these social relations are still relations between positions rather than simply
between individual persons. Of course, the positions are filled by real people, and it
would be absurd to argue that the subjective orientation of the actors within positions is
unimportant for understanding the character of the relations themselves. But to the
extent that these relations are structured systematically, they must be understood as
constituting relations among "empty places" as such.
24. The conventional device of treating organizations as "actors" analogous to human
organisms becomes very difficult to sustain when the organization-environment di-
chotomy is undermined. If the world were really structured as depicted in neoclassi-
al economics - pure competition plus atomistic individuals - then perhaps organiza-
tions could be viewed as organisms interacting with an external environment, but once
organizations interpenetrate each other and social relations cross-cut organizational
boundaries, the metaphor begins to obscure more than it clarifies.
25. See James O'Connor, The Fiscal Crisis of the State (New York: St. Martin's Press,
1973).
26. For an extended discussion of income determination and class relations, see Wright,
Class Structure and Income Determination, chs. 3 and 4; Wright, "Race, Class and
Income Inequality," AJS; and Wright and Luca Perrone, "Marxist Class Categories and
Income Inequality," American Sociological Review 42, 1977.
27. There is a fundamental difference between talking about strata within classes and strata
within society. "Status groups" in the Weberian sense are always conceived as categories
213
within the society at large. Occupational strata in the sense discussed above must be
understood as the structural basis for divisions within classes. Their efficacy in society is
derived from the location within classes, not alongsiMe classes.
28. Sexual divisions also are heavily implicated in divisions within the technical division of
labor. Sex-typing of occupations is if anything a more pervasive feature of capitalist
societies than race- and ethnic-typing. However, such sexual divisions do not create the
same kinds of fractioning of classes as do racial and ethnic divisions, since men and
women workers are bound together within family units. While the linkage between the
occupational structure and the sexual division of labor is certainly important for many
theoretical and political issues, it is of less relevance in the present context.
29. Ronald Aminzade, Class Struggle and Social Change in 19th
Century Toulouse, Ph. D.
Dissertation, Dept. of Sociology, University of Michigan, 1978.
30. For a much more extended discussion of modes of determination, see Wright, Class,
Crisisand the State, ch. 1. For an application of them to the analysis of income inequal-
ity, see Wright, Class Structure and Income Determination, ch. 3.
31. For illustrations of models which include some of these additional elements, see Wright,
Class, Crisis and the State, pp. 27, 165, 223.
32.This is an example where the present discussion is seriously incomplete. While it may be
truethat the class structure places broad limits on the forms of political class formation,
the role of political and ideological factors as selections within those limits may be of
muchmore theoretical interest than the limits themselves. For example, while it may be
truethat the existence of large numbers of people in contradictory class locations may
makethe political formation of the working class as a revolutionary class more difficult,
it is probably the case that the forms of the state and ideology constitute the decisive
factorswhich turn that difficulty into an impossibility. Thus,from the point of view of
socialist strategy, it may be more important to attempt to shift the selection processes
thanexpand the structural limits within which those selections take place.
33. Itshould be noted that this formulation is slightly different from my earlier discussions
ofclass formation and class struggle, in which I argued that class structure directly
limitedclass struggle while class formation simply acted as a selection mechanism.
34.Adam Przeworski, "AHistory of Western European Socialism," Unpublished manus-
cript, Political Science Dept., University of Chicago, 1978.
35.Gosta Esping-Andersen, in Social Democ'racl and Working-Class Politics in the
Modern Wel[are State: Denmark and Suweden, University of Wisconsin-Madison,
Ph.D. thesis, Department of Sociology, 1978, an extremely interesting comparison of
theDanish and Swedish Social Democratic Parties, has argued that the critical differen-
cesin the strategies of the two parties can be explained by the different class structures
whichthey faced. The Danish party, he argues, faced a much larger traditional petty
bourgeoisieand was therefore forced to make a number of important compromises in its
policiesin order to obtain petty bourgeois votes. The Swedish Party, on the other hand,
couldrely more exclusively on a strictly working class electoral base and thus could
pursuea more consistently working class (although reformist) politics. The result of
thesedifferences, according to Esping-Anderson, was that the Danish Party was forced
toadopt policies which ultimately eroded part of its working class
support, whereas the
SwedishParty was able to sustain its working class base in a much more stable manner.
Thisis a good example of how class structure acts as a selection determination of class
struggleswithin limits set by the forms of class organization (reformistworking class
partypolitics).
36.See Michael Burawoy, "Towardsa Marxist Theory of the Labor Process," Politics &
Society,8: 3- 4, 1978; Harry Braverman,Labor and Monopoly Capital (New York:
MonthlyReview Press, 1974);Katherine Stone, "The Origins of Job Structures in the
SteelIndustry," Review
ofRadcal Political Economic's, Summer, 1974; and Steven
Marglin, "What Do Bosses Do?" Review ofRadical Political Economics,Summer, 1974.
214
37. For a discussion of the elaboration of managerial structures as part of a response to class
struggle, see Dan Clawson, Class Struggle and the Rise of Bureaucracy, unpublished
Ph.D. dissertation, Sociology Dept., SUNY, Stonybrook, 1978; Richards Edwards,
Contested Terrain (New York: Basic Books, 1979); and Wright, Class, Crisis and the
State, pp. 67- 71.
38. To my knowledge no really good systematic data exists on changes in the degree of
control of workers over the labor process over time. As should be clear from the
discussion of occupation and class earlier, it is impossible to use occupational statistics
to answer this question. However, some indirect measures, as discussed in Wright and J.
Singlemann, "Proletarianization in Advanced Capitalist Societies," Institute for Re-
search on Poverty Discussion Paper, University of Wisconsin, 1978, do indicate that
within specific sectors of the economy, there seems to be a tendency for autonomy to be
reduced over time.